09:15 <irker870> systemtap: mcermak systemtap.git:refs/heads/master * release-3.0-108-g7a94051 / doc/Tapset_Reference_Guide/publicanize.sh scripts/update-docs: Run publicanize.sh as a part of update-docs. http://tinyurl.com/jhysdsp

11:07 <lorddoskias> is probe end {} execute when ctrl+c is passed on a running stap session?

11:08 <lorddoskias> and if i'm using an arrays what's the correct way to clean them so that no memory leaks are introduced?

11:10 <fche> lorddoskias, yes

11:10 <fche> you don't need to clean anything

11:10 <fche> you don't need to initialize or lock anything

11:10 <fche> stap does it for you.

11:10 <lorddoskias> so probe end { delete global_array } is redundant?

11:10 <fche> yes

11:10 <lorddoskias> cool

12:34 <lorddoskias> fche: where can I get the list of output modifier, since the list at https://sourceware.org/systemtap/langref/Formatted_output.html seems rather empty ?

12:34 <lorddoskias> in particular i'd like to print a pointer so can I use %p ?

12:34 <lorddoskias> or %x?

12:34 <fche> either of those

12:35 <fche> man stap is the master documentation for such things

12:35 <lorddoskias> right thanks :)

12:35 <fche> that langref page shouldn't have empty tables though :)

13:10 <lorddoskias> stack[$return] = sprint_backtrace();

13:10 <lorddoskias> that's a correct usage, no?

13:10 <lorddoskias> since later when i use it such as: printf("%s\n", stack[$dquot]); i get an empty string and just the \n char :(

13:11 <fche> there are some known problems with collecting full backtraces from .return probes, but there should be some content

13:11 <fche> would have to see more of your script to see if what you're doing makes sense

13:18 <lorddoskias> fche: here is it: http://pastebin.com/1t2U2Q4T

13:19 <lorddoskias> everything else works, it's just the printf for the stack that's acting up

13:20 <fche> have you tried printing out the stack[] array, to see if you're getting any good data?

13:20 <fche> stack[$dquot] will be an empty string if stack[$dquot] hasn't been assigned-to before

13:22 <lorddoskias> ah, i finally managed to get some stacktraces

13:22 <lorddoskias> so it seems the code is correct per-se, it's just that it hasn't been assigned for a particular pointer

13:23 <fche> yeah, could be if the dqget calls don't line up one-for-one with remove_tree ones

13:24 <fche> you could do if ($dquot in stack) println(stack[$dquot])

14:25 <lorddoskias> ah, i was wondering whether there is such a "test for membership of array" :)

14:26 <fche> see [man stapref] or [man stap]

15:31 <irker870> systemtap: csanting systemtap.git:refs/heads/csanting/PR11637 * release-3.0-97-gcbd1afb / tapset/uconversions-guru.stp testsuite/systemtap.base/set_user.c testsuite/systemtap.base/set_user.exp testsuite/systemtap.base/set_user.stp testsuite/systemtap.base/set_user_fail.stp: 11637: Added test cases, added to set_user_* functions http://tinyurl.com/z4n6g8o